Use fund data from VG or M*?

[OverseasBH]

Should one use the VG fund data from VG or Morningstar?

With my core bond fund, VBILX/BIV, I keep track of how much is in government securities versus corporates. The problem is that VG and M* report somewhat different percentages for the same portfolio on the same date.

For the Feb 28, VG shows that there is 57.30% of this fund's portfolio in government bonds (Treasuries, agencies, MBS). M* shows that government bonds comprise 62.32% of the fund on the same date.

Both cannot be right but it is hard to know which one to follow, as I assume they both define government bond the same way. Any ideas?

[jebmke]

My guess is that they use data from different dates. The difference Iin your example is immaterial IMO. Personally, I use the data from VG.

[UpperNwGuy]

Morningstar undoubtedly gets its data from Vanguard. I don't trust either one to be always current with the data they display on the internet.

[jebmke]

One way to solve this is to hold two funds in your desired mix. treasury fund and corporate bond fund. Then the mix is under your control.

[dcabler]

One way to solve this is to hold two funds in your desired mix. treasury fund and corporate bond fund. Then the mix is under your control.

Then you get to wonder about whether the treasury fund holds the treasury bonds stated on vanguard or on M*. Same for Corporate bonds.

[sycamore]

Use the average of the two numbers

[nisiprius]

Should one use the VG fund data from VG or Morningstar?

With my core bond fund, VBILX/BIV, I keep track of how much is in government securities versus corporates. The problem is that VG and M* report somewhat different percentages for the same portfolio on the same date.

For the Feb 28, VG shows that there is 57.30% of this fund's portfolio in government bonds (Treasuries, agencies, MBS). M* shows that government bonds comprise 62.32% of the fund on the same date.

Both cannot be right but it is hard to know which one to follow, as I assume they both define government bond the same way. Any ideas?

My ideas are: they are both right, and that their definitions of "government bond" are not the same, and that it doesn't matter anyway...

...and that you should use Vanguard's because (twice a year) they list the bonds in the fund in annual and semiannual report, and in this report they organize the list of bonds according to category. So they are on record as to which are which. If it was really important to you, you could probably find a way to convert that list to an Excel file, add up the numbers, and check the math.

For 12/31/2022, "sovereign bonds, 3.8%" (Japan, Alberta, Indonesia... did you know that BIV held them?), "taxable municipal bonds, 0.4%" (I didn't know there was such a thing), as well as "US Government Securities" and "Agency Bonds and Notes." So you can see how they categorized each bond. Morningstar does have a list of holdings by "sector" but you can't see more than the top ten without a paid subscription.

Why don't I think it matters? Because

• I have no clue whether I think an Indonesian government bond is a "government bond" or not and
• I don't believe there is any meaningful difference between 57.30% and 62.32%.

My suspicion is that in this case the apparent difference is explained by the fact that Vanguard's total is described as U.S. Government Bonds, while Morningstar only says "Government Bonds."

And I have a soapbox rant. People believe financial numbers because they are stated to the dollar, but so many of them depend on arbitrary category definitions, the use of human judgement, and fuzzy boundaries.

Even something as simple as whether a particular stock is "US" or "international" isn't so simple. The last time I checked the S&P 500 included 22 stocks of foreign companies, domiciled in Bermuda, Curaçao, England, Ireland, the Netherlands, and Switzerland. There's a complicated list of rules and someone decided these were "domiciles of convenience" and that Garmin is "really" a U.S. company, not a Swiss company.

Even numbers as basic as "the total return of the S&P 500 for year Y" can be flaky. I once caught two different editions of the same reference book (Ibbotson SBBI Classic Yearbook) giving two different numbers for the same year.

[retiredjg]

For the Feb 28, VG shows that there is 57.30% of this fund's portfolio in government bonds (Treasuries, agencies, MBS). M* shows that government bonds comprise 62.32% of the fund on the same date.

I don't know why a person would be tracking this. Whatever it is, I don't think the difference here is significant enough to worry about. Investing is not very precise. Trying to make it precise will drive you nuts!

[nisiprius]

P.S. I had no idea until ten minutes ago that the Vanguard Intermediate-Term Bond ETF include some foreign bonds.

Why would it do that? I don't know. Vanguard says tracks the Bloomberg U.S. 5–10 Year Government/Credit Float A index. I can't seem to find any Bloomberg factsheet for that index so I don't know if that index includes foreign bonds, or whether Vanguard is throwing in a few of them them to juice the fund's performance... but I will say that would be "unlike Vanguard."

They aren't being sneaky about it:

Web page for BIV

Product summary

• Seeks to track the investment return of the Bloomberg U.S. 5–10 Year Government/Credit Float Adjusted Index, a market-weighted bond index that covers investment-grade bonds with a dollar-weighted average maturity of 5 to 10 years.
• Invests in U.S. government high-quality (investment grade) corporate and investment-grade international dollar-denominated bonds.
• Passively managed, using index sampling.

[retiredjg]

P.S. I had no idea until ten minutes ago that the Vanguard Intermediate-Term Bond ETF include some foreign bonds.

I just stumbled on the same thing while looking at my "US vs International" bond ratio on my Vanguard webpage. Between Total Bond and Intermediate Term Bond, 10% of my bond allocation is international bonds.

I had no idea. I wonder if they have been there all along or if this is a recent development.

[OverseasBH]

Should one use the VG fund data from VG or Morningstar?

With my core bond fund, VBILX/BIV, I keep track of how much is in government securities versus corporates. The problem is that VG and M* report somewhat different percentages for the same portfolio on the same date.

For the Feb 28, VG shows that there is 57.30% of this fund's portfolio in government bonds (Treasuries, agencies, MBS). M* shows that government bonds comprise 62.32% of the fund on the same date.

Both cannot be right but it is hard to know which one to follow, as I assume they both define government bond the same way. Any ideas?

My ideas are: they are both right, and that their definitions of "government bond" are not the same, and that it doesn't matter anyway...

...and that you should use Vanguard's because (twice a year) they list the bonds in the fund in annual and semiannual report, and in this report they organize the list of bonds according to category. So they are on record as to which are which. If it was really important to you, you could probably find a way to convert that list to an Excel file, add up the numbers, and check the math.

For 12/31/2022, "sovereign bonds, 3.8%" (Japan, Alberta, Indonesia... did you know that BIV held them?), "taxable municipal bonds, 0.4%" (I didn't know there was such a thing), as well as "US Government Securities" and "Agency Bonds and Notes." So you can see how they categorized each bond. Morningstar does have a list of holdings by "sector" but you can't see more than the top ten without a paid subscription.

Why don't I think it matters? Because

• I have no clue whether I think an Indonesian government bond is a "government bond" or not and
• I don't believe there is any meaningful difference between 57.30% and 62.32%.

My suspicion is that in this case the apparent difference is explained by the fact that Vanguard's total is described as U.S. Government Bonds, while Morningstar only says "Government Bonds."

And I have a soapbox rant. People believe financial numbers because they are stated to the dollar, but so many of them depend on arbitrary category definitions, the use of human judgement, and fuzzy boundaries.

Even something as simple as whether a particular stock is "US" or "international" isn't so simple. The last time I checked the S&P 500 included 22 stocks of foreign companies, domiciled in Bermuda, Curaçao, England, Ireland, the Netherlands, and Switzerland. There's a complicated list of rules and someone decided these were "domiciles of convenience" and that Garmin is "really" a U.S. company, not a Swiss company.

Even numbers as basic as "the total return of the S&P 500 for year Y" can be flaky. I once caught two different editions of the same reference book (Ibbotson SBBI Classic Yearbook) giving two different numbers for the same year.

Yes, I could not find how M* defined government bonds, but your explanation seems the most likely reason. So I used Portfolio Watch to see if this was true, as it breaks down bonds by US v international. It kind of does, but not really that precisely, so it not possible to say conclusively. With the Portfolio Watch switch from using a fund's objectives to the actual data, I am still not clear if they are taking live data (and so this represents the fund's bond holdings as of March 16) or if they are only taking in the holdings as of the end of February, which might explain this difference.

Totally agree as the fuzziness of boundaries of the financial data.

[jebmke]

Yes, I could not find how M* defined government bonds, but your explanation seems the most likely reason. So I used Portfolio Watch to see if this was true, as it breaks down bonds by US v international. It kind of does, but not really that precisely, so it not possible to say conclusively. With the Portfolio Watch switch from using a fund's objectives to the actual data, I am still not clear if they are taking live data (and so this represents the fund's bond holdings as of March 16) or if they are only taking in the holdings as of the end of February, which might explain this difference.

Totally agree as the fuzziness of boundaries of the financial data.

Would not be surprised if they are using data that is even older - end of last quarter.

[OverseasBH]

P.S. I had no idea until ten minutes ago that the Vanguard Intermediate-Term Bond ETF include some foreign bonds.

Why would it do that? I don't know. Vanguard says tracks the Bloomberg U.S. 5–10 Year Government/Credit Float A index. I can't seem to find any Bloomberg factsheet for that index so I don't know if that index includes foreign bonds, or whether Vanguard is throwing in a few of them them to juice the fund's performance... but I will say that would be "unlike Vanguard."

They aren't being sneaky about it:

Web page for BIV

Product summary

• Seeks to track the investment return of the Bloomberg U.S. 5–10 Year Government/Credit Float Adjusted Index, a market-weighted bond index that covers investment-grade bonds with a dollar-weighted average maturity of 5 to 10 years.
• Invests in U.S. government high-quality (investment grade) corporate and investment-grade international dollar-denominated bonds.
• Passively managed, using index sampling.

I also noted in Portfolio Watch something that they must have just implemented this week (although it states this changed January 2022) that it now classifies bonds as international by their currency instead of by their country of origin. This led my percentage of international bonds to jump 10%!

[OverseasBH]

Yes, I could not find how M* defined government bonds, but your explanation seems the most likely reason. So I used Portfolio Watch to see if this was true, as it breaks down bonds by US v international. It kind of does, but not really that precisely, so it not possible to say conclusively. With the Portfolio Watch switch from using a fund's objectives to the actual data, I am still not clear if they are taking live data (and so this represents the fund's bond holdings as of March 16) or if they are only taking in the holdings as of the end of February, which might explain this difference.

Totally agree as the fuzziness of boundaries of the financial data.

Would not be surprised if they are using data that is even older - end of last quarter.

One of my biggest complaints about Portfolio Watch was that it used a fund's objectives to classify (e.g. government vs. corporate, short vs. intermediate term, etc.) instead of actual data. They apparently have changed that and as they now have published the prior month's portfolios, I would think it would not go back further than that, would it?

[OverseasBH]

P.S. I had no idea until ten minutes ago that the Vanguard Intermediate-Term Bond ETF include some foreign bonds.

I just stumbled on the same thing while looking at my "US vs International" bond ratio on my Vanguard webpage. Between Total Bond and Intermediate Term Bond, 10% of my bond allocation is international bonds.

I had no idea. I wonder if they have been there all along or if this is a recent development.

Further to my comment about the change in Portfolio Watch regarding the classification of international bonds and the 10% increase in my allocation this week, although I made no trades:

Portfolio Watch now determines a bond's region by currency. Portfolio Watch previously determined bond regions by country of origin, so your current analysis may differ from a previous analysis.

[retiredjg]

P.S. I had no idea until ten minutes ago that the Vanguard Intermediate-Term Bond ETF include some foreign bonds.

I just stumbled on the same thing while looking at my "US vs International" bond ratio on my Vanguard webpage. Between Total Bond and Intermediate Term Bond, 10% of my bond allocation is international bonds.

I had no idea. I wonder if they have been there all along or if this is a recent development.

Further to my comment about the change in Portfolio Watch regarding the classification of international bonds and the 10% increase in my allocation this week, although I made no trades:

Portfolio Watch now determines a bond's region by currency. Portfolio Watch previously determined bond regions by country of origin, so your current analysis may differ from a previous analysis.

I suppose that could explain what I just noticed today. I don't recall ever seeing international bonds before and I'm pretty sure I've looked at that graph before.

[OverseasBH]

P.S. I had no idea until ten minutes ago that the Vanguard Intermediate-Term Bond ETF include some foreign bonds.

I just stumbled on the same thing while looking at my "US vs International" bond ratio on my Vanguard webpage. Between Total Bond and Intermediate Term Bond, 10% of my bond allocation is international bonds.

I had no idea. I wonder if they have been there all along or if this is a recent development.

Further to my comment about the change in Portfolio Watch regarding the classification of international bonds and the 10% increase in my allocation this week, although I made no trades:

Portfolio Watch now determines a bond's region by currency. Portfolio Watch previously determined bond regions by country of origin, so your current analysis may differ from a previous analysis.

I suppose that could explain what I just noticed today. I don't recall ever seeing international bonds before and I'm pretty sure I've looked at that graph before.

Is VG generally updating their statistics this week? I noted that Portfolio Watch is now telling me that I am underweight US large caps. It states that the US market for large caps is 78.6%. According to M*, TSMI is only 72% large cap.

Not sure how TSMI can be 72% large cap but I am being flagged that I am underweight the US market large cap percentage.

[nisiprius]

Portfolio Watch. Don't get me started. It's been wildly unreliable all along, and it now has a fresh set of gross bugs. The only stock funds I own are Total Stock and Total International, and it says I have 0% invested in consumer discretionary.

[Geologist]

P.S. I had no idea until ten minutes ago that the Vanguard Intermediate-Term Bond ETF include some foreign bonds.

I just stumbled on the same thing while looking at my "US vs International" bond ratio on my Vanguard webpage. Between Total Bond and Intermediate Term Bond, 10% of my bond allocation is international bonds.

I had no idea. I wonder if they have been there all along or if this is a recent development.

One advantage of obtaining annual reports is you can judge this sort of question. In 2012, the Intermediate-Term Bond Index Fund had 5.5% Sovereign Bonds (US Dollar denominated) and the Total Bond Index Fund had 4.5% Sovereign Bonds.

Consequently, I conclude that they have been there all along.

[Geologist]

For the OP, if you want to track government versus corporate bonds, I wouldn't use summary website totals. I would look at annual and semi-annual reports.

If you really want great detail (and I don't know why because the point of a fund is that the managers deal with this), then you need to look at the complete list of holdings that is filed quarterly with the SEC (and it is probably at vanguard.com somewhere).

[toddthebod]

My suspicion is that in this case the apparent difference is explained by the fact that Vanguard's total is described as U.S. Government Bonds, while Morningstar only says "Government Bonds."

I think that is quite obviously the case if you look at the entire breakdown. Note there is no foreign category in Morningstar's list.

Morningstar:

Sectors Investment %
Government 62.32
Municipal 0.36
Corporate 36.61
Securitized 0.13
Cash & Equivalents 0.57
Other 0.00

Vanguard:

Issuer % of fund
Finance 13.70%
Foreign 4.30%
Industrial 21.10%
Other 0.60%
Treasury/Agency 57.30%
Utilities 3.00%

[OverseasBH]

My suspicion is that in this case the apparent difference is explained by the fact that Vanguard's total is described as U.S. Government Bonds, while Morningstar only says "Government Bonds."

I think that is quite obviously the case if you look at the entire breakdown. Note there is no foreign category in Morningstar's list.

Morningstar:

Sectors Investment %
Government 62.32
Municipal 0.36
Corporate 36.61
Securitized 0.13
Cash & Equivalents 0.57
Other 0.00

Vanguard:

Issuer % of fund
Finance 13.70%
Foreign 4.30%
Industrial 21.10%
Other 0.60%
Treasury/Agency 57.30%
Utilities 3.00%

When looking at this portfolio using both lists, how do you evaluate the credit risk of those foreign bonds?

In round numbers, there are 57% US government bonds, 37% investment grade corporate bonds, and around 5% foreign government bonds.

Without getting into specific security analysis, should one generally lump the foreign government bonds with the US government bonds in terms of credit quality or should they be lumped with the investment grade corporates?

[Geologist]

My suspicion is that in this case the apparent difference is explained by the fact that Vanguard's total is described as U.S. Government Bonds, while Morningstar only says "Government Bonds."

I think that is quite obviously the case if you look at the entire breakdown. Note there is no foreign category in Morningstar's list.

Morningstar:

Sectors Investment %
Government 62.32
Municipal 0.36
Corporate 36.61
Securitized 0.13
Cash & Equivalents 0.57
Other 0.00

Vanguard:

Issuer % of fund
Finance 13.70%
Foreign 4.30%
Industrial 21.10%
Other 0.60%
Treasury/Agency 57.30%
Utilities 3.00%

When looking at this portfolio using both lists, how do you evaluate the credit risk of those foreign bonds?

In round numbers, there are 57% US government bonds, 37% investment grade corporate bonds, and around 5% foreign government bonds.

Without getting into specific security analysis, should one generally lump the foreign government bonds with the US government bonds in terms of credit quality or should they be lumped with the investment grade corporates?

The corporate bonds are not all the same credit quality. They are a whole range. So you can't lump all of them together.

I have investments in the TIAA variable annuity equivalent of the Total Bond Index fund and its annual report, long ago, used to list the bond ratings of each individual bond. However, they stopped doing this many years ago or I could have told you what the bond rating agencies rate sovereign bonds.

[Geologist]

Looking at the 2021 Annual report, the Sovereign Bonds range from provinces of Canada to Uruguay to Mexico to Japan to Korea. If I had to guess, there is a range of credit qualities here.

[toddthebod]

My suspicion is that in this case the apparent difference is explained by the fact that Vanguard's total is described as U.S. Government Bonds, while Morningstar only says "Government Bonds."

I think that is quite obviously the case if you look at the entire breakdown. Note there is no foreign category in Morningstar's list.

Morningstar:

Sectors Investment %
Government 62.32
Municipal 0.36
Corporate 36.61
Securitized 0.13
Cash & Equivalents 0.57
Other 0.00

Vanguard:

Issuer % of fund
Finance 13.70%
Foreign 4.30%
Industrial 21.10%
Other 0.60%
Treasury/Agency 57.30%
Utilities 3.00%

When looking at this portfolio using both lists, how do you evaluate the credit risk of those foreign bonds?

In round numbers, there are 57% US government bonds, 37% investment grade corporate bonds, and around 5% foreign government bonds.

Without getting into specific security analysis, should one generally lump the foreign government bonds with the US government bonds in terms of credit quality or should they be lumped with the investment grade corporates?

There's a separate breakdown of the credit quality. Would you rather invest in Microsoft bonds or the government of Uruguay? It's not all about government versus corporate.

[OverseasBH]

Looking at the 2021 Annual report, the Sovereign Bonds range from provinces of Canada to Uruguay to Mexico to Japan to Korea. If I had to guess, there is a range of credit qualities here.

I guess I am only trying to put these into one of two categories, either government (no credit risk) or investment grade (some credit risk). The US treasury/agency bonds go into the former category and the corporate bonds go into latter category. But do any of the foreign government bonds belong in the government category, if it is defined as having no credit risk?